After receiving institutional review board approval, a retrospective chart review of patients who underwent RSA at Atrium Health Wake Forest Baptist Medical Center between 2015 and 2022 was conducted. Inclusion criteria included (1) RSA with availability of pre- and postoperative radiographic imaging and (2) a minimum 2-year follow-up with completed PROMs. Patients who underwent RSA for a proximal humerus fracture were excluded, as were all RSA revision procedures. All operations were performed by 1 of 4 fellowship-trained orthopaedic surgeons (B.R.G., B.R.W., N.A.T., A.W.R.), using either the Exactech Equinoxe or Tornier Aequalis (Stryker) implant systems. For the purposes of this study, we define “modern implants” as those having a more anatomic neck shaft angle (equal to or less than 145°) and a lateralized glenoid component, achieved by baseplate lateralization, baseplate augment, or glenosphere lateralization. The RSA angle could impact PROMs and ROM by altering the biomechanics of the shoulder, specifically deltoid tensioning and center of rotation, as well as impacting bony impingement.

Data collection was performed via review of the electronic medical record. Age, sex, laterality, limb dominance, body mass index, comorbidities, and indication for surgery were documented for each patient. Pre- and postoperative RSA angles were obtained through measurement on a standardized Grashey view radiograph using the Picture Archiving and Communication System software (Philips IntelliSpace PACS) ( Fig 1 ). The RSA angles were measured by the first author (N.E.D.) and audited by the second author (A.W.R.), a fellowship-trained orthopaedic surgeon. The RSA angle was measured by identifying the angle between a line along the inferior part of the glenoid fossa and a line perpendicular to the floor of the supraspinatus fossa, as defined by Boileau et al. Outliers were defined as those patients who had a postoperative RSA angle greater than 10°.

Measurements of preoperative reverse shoulder arthroplasty (RSA) angle (A) and postoperative RSA angle (B) on Grashey radiographs of the left shoulder of a 61-year-old man. The RSA angle was defined as the angle between the inferior part of the glenoid fossa and the perpendicular line to the floor of the supraspinatus fossa.

All PROM data were collected either through the Surgical Outcomes System (Arthrex), electronically or on paper prior to clinic visits, or by telephone administration. Specifically, the American Shoulder and Elbow Surgeons (ASES) score, visual analog scale (VAS), and Single Assessment Numeric Evaluation (SANE) surveys were administered. If the patient’s chart did not contain 2-year postoperative PROM data, the patient was called, and the PROMs were administered via a virtual telemedicine visit. An ASES score greater than or equal to 80 is considered excellent. Additionally, the minimal clinically important difference (MCID) was calculated using a distribution-based approach (i.e., half standard deviation of the delta in scores from preoperative to postoperative time points) for ASES, VAS, and SANE scores. The frequency at which patients achieved MCID for each respective PROM measure was also recorded.

Pre- and postoperative ROM was also obtained at a minimum of 1 year postoperatively, which consisted of forward flexion, external rotation, and internal rotation. ROM was visually measured by 1 of the 4 fellowship-trained orthopaedic surgeons listed and recorded. These data were extracted from the 1-year follow-up note in each patient’s electronic medical record. Forward flexion and external rotation at 0° of abduction were measured and recorded in degrees; internal rotation in adduction was quantified using a numeric scale as described by Levy et al. In this system, internal rotation measurements with the arm in adduction were translated into a 10-point scale. Patients received 2 points if they could actively rotate internally to reach the buttock or greater trochanter. Additional points were allocated based on the highest level reached during active internal rotation: sacrum to L4 earned 4 points; L3 to L1, 6 points; T12 to T8, 8 points; and T7 to T1, 10 points. Easily identifiable anatomic landmarks, such as the posterior superior iliac spine, iliac crest, and inferior angle of the scapula, were used to approximate the level of the spine reached by the patient.

Postoperative complications were recorded, including pain, instability, impingement, fracture, and the need for revision RSA. Pain was considered a complication if the patient had unscheduled return visits with ongoing pain as the primary concern and no other etiology identified.

All data were screened to determine if they met all parametric statistical assumptions prior to analysis. Both Pearson and Spearman correlation analyses were performed to determine whether there were any linear or monotonic relationships between postoperative RSA angle and the other variables. Analysis was conducted using SPSS software (version 29.0.2; IBM Corp). Averages are reported as mean ± standard deviation unless stated otherwise.